Characterization of T-cell receptor repertoire and correlation with tumor mutational landscape in lung cancer.

Abstract

2577 Background: T-cell receptor (TCR) repertoire represents an overall immune condition which is closely related to pathogens- or tumor-associated T cell responses. Increasing evidence suggests that TCR repertoire is expected to undergo cancer-specific changes during tumorigenesis, supporting that TCR characteristics can serve as novel markers to indicate early cancer progression. Here, we performed a systematic T cell repertoire analysis in lung cancer. We identified cancer-enriched TCR signatures and constructed a comprehensive lung cancer database integrating TCR diversity, tumor mutational profiles, immune markers expression and HLA genotypes. Methods: A total of 988 tissue samples and 3,360 blood samples were obtained from patients with lung cancer, in addition to 2,699 blood samples from healthy individuals. Multiplex-PCR-based sequencing of the CDR3 regions of TCR-β chains was applied to the tissue and blood samples. TCR repertoires were analyzed using MiXCR and VDJtools. Whole exome sequencing was performed on partial tumor tissues to profile the mutational landscape and HLA genotypes. An in-house pipeline, considering the frequency and distribution of CDR3, was utilized to identify cancer-enriched CDR3 sequences. The PD-L1 expression was evaluated and quantified using CPS. Furthermore, an enrich score was developed to measure the content of cancer-related TCRs by aligning against the built cancer-enriched CDR3 dataset. Spearman's rank correlation was used to assess relationships between variables. Results: Lung cancer exhibited significantly lower Shannon index, Simpson index and evenness index in TCR repertoire (p<0.001), both in tissue and blood samples, when compared with healthy blood samples. Approximate 3% T-cell clonotypes within lung tissues were detected in blood TCR repertoires. A correlation was observed between the decreased TCR diversity and a higher tumor mutational burden (TMB)(r = -0.15, p < 0.01), as well as a higher variant allele frequency (VAF) (r = -0.19, p < 0.01). In this study, 3,652 and 3,840 CDR3 sequences were identified to be enriched in lung cancer tissue and blood respectively. A significant difference in cancer-enriched score was observed between cancer and healthy TCR repertoires (p < 0.001). The cancer-enriched score showed a significant positive correlation with TMB, particularly in TP53- mutant tumors (r = 0.27, p < 0.001). A higher CPS (PD-L1) was associated with a higher cancer-enriched score (r = 0.18 ,p < 0.01). Conclusions: Cancer-related T-cell clonotypes in the tumor microenvironment and peripheral blood informs the changes in anti-tumor responses, supporting the potential application of TCR signatures in early cancer detection. This study identified the lung cancer-related TCR signatures and provided a comprehensive lung cancer TCR database, revealing the relationship between TCR characteristics and mutational profiles. Clinical trial information: ChiCTR2200055761.